Unable to renewsliver or get sliverstatus at internet2 aggregate for active slivers exchanging traffic

[lnevers@bbn.com]

Test scenario includes two stitched slices (ln-ext1 and ln-ext2). The slices include:

• Slice ln-ext1 -> 1VM@ Utah IG and 1 VM @ GPO IG (endpoing are pinging each other)
• Slice ln-ext2 -> 1VM@ Utah IG and 1 VM @ GPO IG (endpoing are pinging each other)

Both slices are concurrently active and pings are running as the failures are captured below.

Issued a renewsliver for both slivers, which both failed at Internet2 aggregate:

For Slice ln-ext1-> renew sliver successful@ Utah IG, Utah PG and GPO IG, but fails at Internet2 with the following error:

WARNING:omni:Failed to renew sliver urn:publicid:IDN+ch.geni.net:ln-prj+slice+ln-ext1 
on unspecified_AM_URN+ (http://geni-am.net.internet2.edu:12346) (got result 'None'). 
Error from Aggregate: code 12. sfa AM code: 12: : RenewSliver: Non existing record 
urn:publicid:IDN+ch.geni.net:ln-prj+slice+ln-ext1, .

For Slice ln-ext2-> renew sliver successful@ Utah IG, Utah PG and GPO IG, but fails at Internet2 with the following error:

WARNING:omni:Failed to renew sliver urn:publicid:IDN+ch.geni.net:ln-prj+slice+ln-ext2 
on unspecified_AM_URN+ (http://geni-am.net.internet2.edu:12346) (got result 'None'). 
Error from Aggregate: code 5. sfa AM code: 5: : RenewSliver: Internal API error: 
No active RSpecRunnner thread 'urn:publicid:IDN+ch.geni.net:ln-prj+slice+ln-ext2' 
to excute renewal!.

A sliverstatus for Slice ln-ext1 shows:

Error from Aggregate: code 12. sfa AM code: 12: : SliverStatus: Non existing record 
urn:publicid:IDN+ch.geni.net:ln-prj+slice+ln-ext1, .

A sliverstatus for Slice ln-ext2 shows:

Slice urn:publicid:IDN+ch.geni.net:ln-prj+slice+ln-ext2 at AM http://geni-am.net.internet2.edu:12346 
has overall SliverStatus: failed.

Traffic is still successfully being exchanged with the above failures

[lnevers@bbn.com]

In this scenario the ln-exp1 sliver had been shutdown as part of an earlier partially successful shutdown operation, so maybe the ln-ext1 results are appropriate.

Slice ln-ext2 should be in ready state, as no operations were run after the initial createsliver and before the shutdown operation captured in the ticket description.

[xyang@maxgigapop.net]

IOM AM has been down and up couple of times for some updates.

Please rerun the scenario to see if replicable.

[lnevers@bbn.com]

Recreated both ln-ext1 and ln-ext2 slices and was able to renew slivers and get sliver status for both without any problem. The issues seen in the initial description must have been related to the ION restart and updates. Closing ticket.